I've been at it for a few yrs now. If you want to mine BTC directly you'll have to go w/an ASIC miner. The biggest player is Bitmain as they deliver a product relatively on time and are generally top of the chain. The problem w/ASICS is they turn into door stops as they become obsolete and will only mine a single algorithm. Obsolescence is a matter of months mind you. A rig that cost you 3k will be worth 100$ almost a year and a half later. I would say buy an obsolete one for pennies on the dollar to get acquainted with setup and see if you want to deal w/the heat and noise (you can pick one up for less than 50$). These machines are by far the simplest to setup. https://www.asicminervalue.com/ You can mine bitcoin indirectly by using other algos using CPU/GPU and cashing into bitcoin. This is where something like NiceHash comes into play. While the immediate returns of GPUs are not as high as ASIC, the GPUs are much more flexible in mining other algos. They aren't competitive to mine BTC directly but allowing reprogramming for other algos. Basically, your GPU rig will mine Ethereum/shitcoinsxyz, etc... and trade into BTC. Nicehash does this automatically for you and you don't have to use an exchange. I'm fairly certain Nicehash also sells hashing power (among other comps) so you don't have to physically deal w/the equipment yourself. You can do the above and keep the ethereum or shitcoin of your flavor and then cash out into BTC using an exchange but it's an extra layer. It's feared that GPU rigs are mostly profitable due to ethereum at the moment and will become unprofitable once ETH goes to proof of stake, so beware. I won't get into FPGAs as that's a whole 'nother can of worms not worth opening.
This is great to read (from an ex-bitcoin miner)! Thanks for posting it. Honestly, if I can get free or next to free electricity as the OP mentioned, I'd setup a GPU mining rig (for altcoins) and get some passive income or to stack some sats. It's a fun hobby and keeps you abreast of technical innovation in the blockchain space
Just out of curiosity, what do the returns look like? How much of your time does it require? Can you scale up as a one man operation? How much can BTC price move without things turning unprofitable? No worries if you don’t feel like answering any of it...just find the whole interesting and know nothing.
I could of been a miner in 2010 and then some idiot on bitcointalk was bitching about how it burned out his mother board. So my three day jab at it ended. (I didn't want to fry my $2800 Dell laptop. Then I let a 'friend' who was uninformed tell me "it's a scam don't invest money in it." My dad who was dirt poor said "Why wouldn't you just use paypal?". Morale of the story don't listen to doubters. Do whatever the hell you want as that will probably turn out to be the right decision for you. All of this said I let the idea that I missed it at pennies stop me from buying it many times on the way up. When I did finally start buying it in the $180's I ended up gambling it and using it to pay back debts to people who also didn't keep it. Ultimately I am over here with a fraction of a coin and still a believer. It's nice to think about what could have been if I DIDN'T ask for anyone's opinion and just said fuck it.. What's the worst that can happen if I buy 100k of these and they end up worthless? (At the time BTC was swinging between 0.008 and 0.32.) (I've wasted far more in actual dollars on women who came and went. )
returns are pretty low for me atm as I'm running 3-4 yo hardware (GPUs and FPGA). So maybe 5/day/rig. Top of the line ASICs are doing 30/day or so atm. This after subtracting pwr. I'm planning to move across the country so haven't bothered upsizing. Time was substantial when I started as the amount of dedicated hardware and software was lower than it is now so there were uptime issues all the time. Now that there's dedicated hardware and software out, maybe an hr a day at most. Keep in mind this is as much a hobby as it is a passive income stream, so I don't stress too much if I have a rig go down. Now a days you have email notifications, remote monitoring and rebooting thru your phone, and hardware watchdogs so it's easier to keep rigs online. ASICs are pretty fool proof so manpower's negligible unless you get a lemon. It's definitely scalable for a one man team, it's what's so great about it. The limiting factor is space, money, infrastructure (power), and know-how. From what I've learned, it takes me less than an hr to build a rig and have it up and running. ASICS, maybe 10-15 mins. The last question's the wild card honestly and difficult to answer. It depends on how much you went in at, if you overpaid for hw, if you didn't fuck up buying the wrong hw, and obviously the price volatility. One thing that was a common fuck up for instance was buying an ASIC and then a competitor releasing theirs w/double the hashrate making your machine a brick. It used to be also that manufacturers would have pre-sales and then not deliver on time or another comp. would release before they did making your purchase a bad one. To avoid that, you have to buy from the top dogs and reputable vendors w/a history of delivering at the promised day/week Everything that's decent's coming out of China so there's speculation that they're always building and running something better and maybe even selling you rigs they've been running for months and giving them a shine before boxing them up. The likelier scenario is they're running upscale custom machines themselves for private farms. What I've noticed is they're building them w/razor thin specs, in essence guaranteeing you'll be back in 6 months for their shinier model. It's incredibly wasteful tbh.
Very interesting, thanks for taking the time with that. Also, I apologize for being lazy with a few of those questions as I see some of that data provided directly in the link you shared. I suppose one could hedge out the price risk using derivatives and diversify away most of the obsolescence risk by slowly building out a farm. That said, with 240 days to breakeven on the top model on that website, it would be scary having 100+ machines running. One big innovation on the hardware side and the whole thing values near zero. I'd imagine there's some massive farms out there though.
Isn't there an implied diminishing return for mining where energy costs grow, the ledger size and complexity grows even faster, yet fewer and fewer coins are available to be mined? If true, the value of BTC has to (must) inflate to compensate miners, otherwise the whole thing would crash. BTC has the same pyramid scheme that fiat currency has, but fiat is dependent on a government for stability and BTC is dependent on a consortium of foreign miners AND value inflation. I still prefer the former, but am happy for those who have capitalized off the latter.
Very good question. The economics of bitcoin btc mining is self-adjusting approx ~10days, when the bitcoin network adjusts the difficulty of the "puzzle" with the built-in rules and algorithm as specified in the bitcoin protocol. There are events when the change in economics is drastic and thus, the network hash rate may suddenly increase or decrease, which in turn becomes a key factor of the difficulty for solving the future blocks An example would be 8 months ago, when the price of bitcoin was not as high as today, the rewards for each block got cut in half, imagine you as a miner, your salary has been cut in half, what do you do? Many miners turned off older, less efficient hardware that were unprofitable (they mine less bitcoins than electricity costs). I don't follow it, but I believe the bitcoin network hash rate dropped. When the difficulty got low enough, some of the miners started turning on the old hardware again, thus increasing the network hash rate and difficulty. You can imagine that his is all decentralized and no one calls out a zoom meeting for all the miners to meet and agree on stuff. The economics for each miner is independent, some have free electricity or very low costs and would never turn off the miners except to make room for new ones. All things above also relate to the price. As the price increases, new miners join and network hash rate increases and difficulty increases. If the bitcoin price crashes in half in a few days, it will be like an unscheduled bitcoin halving event for the miners. Yes, 100 years from now, there won't be any more bitcoins to mine and the miners will have to live with the transaction costs, which may be several bitcoins if a lot of transactions occur on the bitcoin network and if the price of each bitcoin is $1M, then there will still be a lot of miners and network hash rate and difficulty will still be very high, but instead of looking 100 years from now, you can always see the progress every 10 days or every 4 years bitcoin halving event.